This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: [scev] Re: Strange IV choices?
- From: Richard Guenther <rguenth at tat dot physik dot uni-tuebingen dot de>
- To: gcc at gcc dot gnu dot org
- Date: Tue, 14 Dec 2004 16:37:48 +0100 (CET)
- Subject: Re: [scev] Re: Strange IV choices?
On Tue, 14 Dec 2004, Richard Guenther wrote:
> On Tue, 14 Dec 2004, Steven Bosscher wrote:
>
> > On Dec 14, 2004 04:02 PM, Richard Guenther <rguenth@tat.physik.uni-tuebingen.de> wrote:
> > > No, but does this help the scev analyzer with currently
> > > unknown evolution?
>
> It seems to depend on the level of "indirection" in the C++ class
> hierarchy, if I short-cut one level, it works again. Are there
> any complexity limits in the scev analyzer? How would they depend
> on the amount of "local" variables?
>
> I'm slightly confused. Maybe someone can have a look at the ivopts dump.
"Testcase" (sorry, big) at
http://www.tat.physik.uni-tuebingen.de/~rguenth/gcc/perf.iu.ii.gz
Richard.
--
Richard Guenther <richard dot guenther at uni-tuebingen dot de>
WWW: http://www.tat.physik.uni-tuebingen.de/~rguenth/